Best prompts for research before buying

AI prompts for research before buying, from cars and tech to stocks. Add your budget and criteria and get a shortlist you can verify.

Big purchases and investment picks both go wrong when you only compare the price and skip the rest. These prompts turn your budget and must-haves into a comparison table and a verification checklist before you sign at a dealer, click checkout online, or buy a stock or ETF.

5 prompts

AI prompt to research a new car before buying

For narrowing new or used cars before a dealer visit. In order to visit the car dealer ship without having to worry getting ripped off.

Prompt

Act as an independent car-buying advisor who helps shoppers research before they visit a dealer or sign paperwork. Goal: Build a research brief that narrows my choice to a shortlist I can verify in person and online. Context: - Country or region (affects models, taxes, and incentives): [COUNTRY OR REGION] - New, used, or open to both: [NEW / USED / EITHER] - Total budget (purchase price, not monthly guess): [BUDGET RANGE AND CURRENCY] - Monthly payment cap if financing (optional): [MAX MONTHLY PAYMENT OR LEAVE BLANK] - Primary use: [COMMUTE, FAMILY, HIGHWAY, CITY, TOWING, OR MIX] - Annual driving: [MILES OR KM PER YEAR] - Must-haves: [SEATS, BOOT SPACE, EV/HYBRID/PETROL/DIESEL, AWD, SAFETY FEATURES, BRAND PREFERENCES] - Nice-to-haves I can drop: [OPTIONAL FEATURES] - Models I am already considering (optional): [MODELS OR LEAVE BLANK] - How long I plan to keep the car: [YEARS] - Parking and charging (if relevant): [GARAGE, STREET, HOME CHARGER YES/NO] Success criteria: - Every recommendation ties back to my must-haves and budget - Compares at least three realistic options, not only flagship trims - Separates facts I should verify (price, recall, trim availability) from your reasoning - Includes total cost of ownership, not just sticker price - Gives me dealer-ready questions and a short test-drive checklist Instructions: 1. List up to 5 clarifying questions if my inputs are too thin. If you can proceed, say "Proceeding with assumptions" and list them. 2. State my non-negotiables in one short bullet list pulled from my inputs. 3. Propose a shortlist of 3 to 5 vehicles (make, model, and typical trim level). For each, say why it fits and one reason it might not. 4. Build a comparison table. Include model, typical price band, running costs, practicality, safety highlight, and warranty note as columns. 5. Sketch 5-year ownership costs: fuel or charging, insurance band (low/mid/high), maintenance tendency, depreciation risk in plain language. 6. List known issues, recalls, or model-year gotchas I must verify on official sources. Label each as "verify before buying." 7. Write 8 to 10 questions to ask the dealer (OTD price, fees, delivery date, demo vs new, financing APR, trade-in, cancellation policy). 8. Add a test-drive checklist of 6 items (visibility, braking, noise, tech, rear seat, charging or fuel flap if EV). 9. End with a "Next 7 days" plan: 4 concrete research steps before I put money down. Output format: Return sections 1 through 9 in order. Use the numbered headings exactly as above.

How to use

Fill [MUST-HAVES] with deal-breakers only. Move preferences to [NICE-TO-HAVES] so the shortlist stays wide enough.
Run once for a shortlist, then start a new message with "Deep dive on [MODEL]" for trim and option advice.
Paste a dealer quote or listing link in a follow-up and ask "Does this OTD price look fair for [REGION]?"

Tips

Monthly payment math hides total cost. Decide [BUDGET RANGE] on full price first, then check financing.
Model-year changes matter. Ask for the specific year and facelift when comparing reviews.
Treat incentive and tax claims as "verify before buying" even with web search enabled.

For best results, give your AI access to:web search

Example output

Section 4

| Model | Typical price band | Fuel note | Practicality | Safety highlight | Warranty |
| --- | --- | --- | --- | --- | --- |
| Toyota Corolla Hybrid | €28-32k new | Low fuel use in city | Compact boot | Strong crash scores in EU tests | 3y manufacturer |
| VW Golf 1.5 TSI | €30-35k new | Petrol, higher motorway use | Mid boot | Good adult protection | 2y + service packs vary |

Section 7

- What is the out-the-door price for this VIN, including delivery and registration?
- Which fees are mandatory vs optional on the contract?
- ...

Next 7 days
- Book two test drives on the same day to compare back-to-back
- Check national recall database for each shortlisted VIN year
- ...

AI prompt to research tech before buying online

For tech bought online, for example phones or laptops. You get review signals to trust or discard, and understand the different options available on the market.

Prompt

Act as a tech shopping advisor focused on online purchases. Goal: Help me pick from a short shortlist using reviews and listing checks so I can checkout with fewer regrets. Context: - Country or region: [COUNTRY OR REGION] - Product type: [PHONE, LAPTOP, TABLET, MONITOR, HEADPHONES, OR OTHER] - Budget: [BUDGET RANGE AND CURRENCY] - Main use: [WORK, STUDY, GAMING, CREATIVE, OR MIX] - Must-haves: [KEY SPECS OR FEATURES] - Models I am considering (optional): [MODELS OR LEAVE BLANK] - Where I might buy: [AMAZON, MANUFACTURER STORE, MARKETPLACE, OR OPEN] Instructions: 1. Ask up to 3 clarifying questions if needed. Otherwise say "Proceeding with assumptions" and list them. 2. Suggest 3 products that fit my must-haves and budget. For each: one fit reason and one drawback. 3. Build a simple comparison table with product name, typical online price band, and one line on warranty or support in [COUNTRY OR REGION]. 4. Estimate true checkout cost: item price plus tax, shipping, and any likely accessories. Flag subscription upsells if relevant. 5. Review research for my shortlist: - Name 2 to 4 source types worth checking for [PRODUCT TYPE] (long-form tests, retailer reviews, forums, etc.) - Summarize recurring pros and cons from recent reviews, filtered for [MAIN USE] - List fake or manipulated review signals: sudden 1-star burst from new accounts, vague non-specific complaints with no photo or detail, complaints that describe a different model or use case, and reviews that read more like a competitor ad than a user report - Give 3 search phrases I should use to find real owner complaints, not just launch hype 6. List listing red flags on the store page (wrong model year, grey import, missing warranty, fake discounts, bundled junk). Mark each "verify before buying." 7. Give 5 questions to check on the product page or with seller chat before I pay (returns, restocking fee, warranty start, new vs refurb, who handles RMA). 8. End with a "Before checkout" checklist of 5 steps I can do in one sitting (include reading reviews and checking the listing). Output format: Return sections 1 through 8 in order. Use the numbered headings exactly as above.

How to use

Paste a product URL in a follow-up and ask "Review this listing for red flags."
After the shortlist, ask "Dig into owner reviews for [MODEL] focused on [MAIN USE]."

Tips

SKU means Stock keeping unit, it's the specific variant you are buying (model, storage size, color). Reviews for the product family can include a different SKU with different problems. Filter or search by your exact variant.
Star averages lie. Sort by recent and read the reviews with 1 to 3 stars for your exact SKU.
A burst of vague 1-star reviews from new accounts right after a product launch is a competitor signal, not a reliability signal.
Real negative reviews usually name a specific defect or describe follow-up with support.
Compare the same SKU across two sites. Bundles often hide a worse base model.

For best results, give your AI access to:web search

Example output

Section 5

Sources to check
- Two recent long-form reviews that test battery and thermals under load
- Retailer reviews sorted by "most recent" and 3-star to see mixed experiences

Recurring themes (writing use)
- Pros: quiet fans, strong keyboard, long idle battery
- Cons: screen glare, occasional coil whine on some units

Possible fake negatives to discard
- Five 1-star reviews posted the same week from accounts with no other history: likely competitor campaign
- "Broke after one day" with no photo or follow-up and the complaint applies to a different model: skip

Search phrases
- "ThinkPad X1 Carbon Gen 12 coil whine"
- "MacBook Air M3 real world battery writing"

Before checkout
- Read at least five recent 3-star reviews on the exact variant (model, storage, color) I am buying
- Screenshot price and return policy today
- ...

AI prompt to analyze a stock or ETF before investing

Before you buy a stock or ETF. You get fundamentals and risks plus a checklist of what to verify in official filings.

Prompt

Act as an independent investment research assistant. You help me think through a stock or ETF before I commit money. Goal: Build a research brief that shows what this asset is, whether it fits my stated goals, and what I must verify on primary sources before I invest. Context: - Ticker or fund name: [TICKER OR FUND NAME] - Asset type: [STOCK / ETF / BOTH IF UNSURE] - Exchange and country (if known): [EXCHANGE AND COUNTRY OR LEAVE BLANK] - Investment amount I am considering: [AMOUNT AND CURRENCY] - Time horizon: [YEARS OR MONTHS] - Goal: [GROWTH, INCOME, CAPITAL PRESERVATION, SPECULATION, OR MIX] - Risk tolerance: [LOW / MEDIUM / HIGH] - Experience level: [BEGINNER / INTERMEDIATE / ADVANCED] - Existing portfolio context (optional): [WHAT YOU ALREADY HOLD OR LEAVE BLANK] - Why this asset caught my attention: [NEWS, RECOMMENDATION, THEME, OR LEAVE BLANK] - Must-haves: [DIVIDEND, LOW FEES, ESG, SECTOR, GEOGRAPHY, OR OTHER] - Deal-breakers: [HIGH DEBT, CONCENTRATED HOLDINGS, COUNTRY RISK, OR LEAVE BLANK] Success criteria: - Explains what the company or fund actually holds in plain language - Separates verified facts from your reasoning and labels gaps as "verify before investing" - Covers key risks, fees, and fit for my time horizon and risk tolerance - Compares at least one alternative I could consider instead - Ends with a checklist of primary sources I should read myself - States clearly that this is research support, not a buy or sell recommendation Instructions: 1. List up to 5 clarifying questions if my inputs are too thin. If you can proceed, say "Proceeding with assumptions" and list them. 2. Confirm the asset identity: full name, ticker, exchange, asset type, and what business or index it tracks. 3. Summarize in plain language what I would own if I bought this (business model for a stock, top holdings and sector mix for an ETF). 4. Cover fundamentals at a level that matches my experience. For stocks: revenue trend, profitability, balance sheet health, valuation context, and competitive position. For ETFs: expense ratio, tracking method, top holdings, concentration, and dividend policy if any. 5. List the main risks and recent headwinds. Include company-specific, sector, macro, currency, and regulatory risks where relevant. 6. Assess fit for my goal and risk profile over my time horizon. Say when it would still be a bad fit for me, even if the asset itself is solid. 7. Compare one or two alternatives (another ETF in the same space, a broad index fund, or a direct competitor). Note tradeoffs in fees vs diversification. 8. Build a "verify before investing" list of 8 to 10 items I should check on official filings, fund factsheets, exchange data, or recent earnings materials. Do not treat forum posts or social media as sources. 9. Write 5 questions I should answer for myself before buying (position size, worst-case loss, exit plan, tax treatment, overlap with what I already hold). 10. End with a short verdict framed as research, not advice: "Looks worth deeper research if..." and "Probably skip if..." based on my inputs. Output format: Return sections 1 through 10 in order.

How to use

Start with the exact ticker and exchange if you know it. Wrong identity typo is classic mistake.
Run once for the overview, then start a new message with "Deep dive on [TICKER] earnings and balance sheet" for one company.
Paste a fund factsheet or annual report excerpt in a follow-up and ask "What in this filing should change my view?"

Tips

Single-stock bets need a clear position size and exit plan before you research further.
ETF names can look similar. Match ticker and ISIN on the issuer site, not only the marketing title.
Past performance in a web summary is not a forecast. Check fees, holdings, and tracking error instead.

For best results, give your AI access to:web search

Example output

Section 3

Vanguard S&P 500 ETF (VOO) tracks the S&P 500 index. You own a slice of roughly 500 large U.S. companies weighted by market cap, not equal weight.

Section 4

Expense ratio about 0.03%. Top holdings lean heavily to mega-cap tech. Dividends paid quarterly. Tracks the index closely via physical replication.

Section 8

- Confirm current expense ratio on the issuer factsheet (verify before investing)
- Read the latest semi-annual report for tracking difference vs the index (verify before investing)
- Check your broker's currency conversion and withholding tax rules (verify before investing)

Section 10

Looks worth deeper research if you want broad U.S. large-cap exposure for 10+ years and can tolerate drawdowns.

Probably skip if you need capital within two years, want no U.S. concentration, or already hold a nearly identical index fund.

This brief is research support only, not financial advice.

AI prompt to analyze terms and conditions or a privacy policy

When you subscribe to a paid app or service, paste the terms or privacy policy and see what you are agreeing to.

Prompt

Act as a document analyst. Read the document pasted below and explain what I am agreeing to before I click accept or sign. Return: 1. Document identity: company name and title, plus last updated date if shown and what kind of document this is 2. Plain-language summary in one paragraph 3. My obligations: fees and account rules, plus acceptable use and anything I must do to keep access 4. Company powers: how the company uses data and runs the account. Flag broad content rights. Note one-sided term changes and mandatory arbitration if present 5. Privacy details if the paste includes a privacy policy. Cover collection purpose, retention, sharing, plus deletion and export options 6. Watch closely: 5 to 10 clauses worth a second read. For each, cite the section and explain why it matters. Rate how serious each clause is on a low-to-high scale 7. Gaps: vague language or references to documents not included in the paste 8. Questions to ask before accepting 9. Reasonable to accept if... and Get professional advice before accepting if... --- START DOCUMENT --- [PASTE FULL DOCUMENT HERE] --- END DOCUMENT ---

How to use

Paste the full text between the START and END markers. That is all you need to do.
If the site splits terms and privacy into separate pages, paste both into one run.
For one clause only, paste just that section and ask "Explain this in plain language."

Tips

Copy the version shown on the site today.
Search the paste for "sole discretion" and "binding arbitration", especially where changes happen without notice.
When the service handles regulated or sensitive personal data, treat the summary as a first pass and confirm with a lawyer.

Example output

Section 2

By using the app you grant a license for uploaded content and agree to binding arbitration in Delaware. The company can also change the terms by posting an update on the website.

Section 6

Section 8.2 "Changes to service" (high): The company may modify or discontinue features without notice.

Section 8

- Do you sell or share personal data with advertisers, and can I opt out?
- What notice do you give before material term changes?

Section 9

Reasonable to accept if you use a free tier and can leave without penalty.

Get professional advice before accepting if you handle customer data or sign a paid annual contract.

This briefing is reading support only, not legal advice.

AI prompt to analyze a legal document under GDPR

When a purchase includes a data processing agreement, paste it and check GDPR gaps while you are still evaluating the vendor.

Prompt

Act as a GDPR document analyst. Read the legal document pasted below and explain what it means for personal data handling under the EU General Data Protection Regulation. Infer the document type, parties, and jurisdiction from the text. Work only from what is pasted. This is reading support, not legal advice. Return: 1. Document identity: title, parties, last updated date if shown, and what kind of document this is, for example a privacy notice or data processing agreement 2. Plain-language summary in one paragraph 3. GDPR roles: who acts as controller or processor, plus any joint controller or sub-processor roles named 4. Data covered: categories of personal data and purposes, plus lawful bases cited or missing 5. Data subject rights: how access, deletion, portability, and objection are handled, plus any stated response deadlines and contact point 6. Transfers and retention: cross-border transfer mechanism if any, plus retention periods stated or absent 7. Security and breaches: measures described and breach notification commitments 8. Watch closely: 5 to 10 clauses with GDPR risk. For each, cite the section and explain the gap or concern. Rate how serious each clause is on a low-to-high scale 9. Questions to ask before signing or approving 10. Reasonable to proceed if... and Get professional advice before proceeding if... --- START DOCUMENT --- [PASTE FULL DOCUMENT HERE] --- END DOCUMENT ---

How to use

Paste the full text between the START and END markers. That is all you need to do.
For a data processing agreement, include sub-processor lists or security annexes if they are separate files.
Follow up with "Check this against Article 28 processor requirements" for a DPA-only pass.

Tips

Know whether you are the controller or the processor before you act on the summary. The same clause reads differently from each side.
Search the paste for "legitimate interest", "sub-processor", and "Standard Contractual Clauses".
Vague purpose wording or a missing lawful basis is a common gap in privacy notices.

For best results, give your AI access to:web search

Example output

Section 2

This DPA governs how the vendor processes customer contact data on your behalf for support ticketing. It lists sub-processors, requires encryption in transit, and promises deletion within 30 days after contract end.

Section 8

Section 4.2 "Sub-processors" (high): The vendor may appoint new sub-processors with 14 days' notice but no right for you to object.

Section 9

- What lawful basis do you rely on when you process our end-user data?
- Where is data stored, and which transfer tool applies?

Section 10

Reasonable to proceed if you need a standard SaaS DPA and the sub-processor list matches your risk tolerance.

Get professional advice before proceeding if you process special category data or face regulator scrutiny in your sector.

This briefing is reading support only, not legal advice.

Related collections

Best prompts for business leaders
AI prompts for business leaders covering strategy and meetings, plus day-to-day operations. Copy prompts built for planning and daily leadership.
Best prompts for a one-person brand weekend
AI prompts for the weekend a one-person brand leaves the notebook and enters the world. The idea finds a face and meets its first real test.
Best prompts for a weirdly useful Sunday with ChatGPT
ChatGPT prompts for a useful Sunday project you can actually finish. Copy one prompt and get a clear task, whether you juggle or prep for a dealer visit.